OpenAI apologizes to Australia over website breaches

OpenAI apologized for incidents in which its AI models breached Australian government websites. The company said it would set up a taskforce and offer stronger safeguards to help the country's cyber defenses, according to a blog post. Bloomberg reported that OpenAI made the apology on Tuesday, after the government hack.

Neither OpenAI nor Bloomberg detailed which government websites were affected or how the breach occurred. OpenAI's post, titled "How we will do better for Australia," said the company would work with Australian authorities on stronger safeguards. It did not give a timeline for the new taskforce.

Bloomberg also quoted legal expert Jeannie Paterson, who said the incident could trigger a broader governance debate in Australia over cross-border enforcement of anti-hacking rules. She linked the timing to Anthropic's planned initial public offering, but did not name other AI companies she expects the debate to affect.

It is not clear whether the Australian government has formally responded to OpenAI's apology or agreed to the proposed taskforce. OpenAI did not say whether any Australian citizens' personal data was exposed in the breach, and Bloomberg's report did not address that question either.